Research map / Bandits and online learning / Online convex optimization and games
Fast and Furious Learning in Zero-Sum Games: Vanishing Regret with Non-Vanishing Step Sizes
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use Gradient Descent with a fixed step size | Online learning in 2x2 zero-sum games with a unique fully mixed Nash equilibrium · 2x2 zero-sum games | regret: O(\eta\cdot T) -> O(\sqrt{T}) | Helped |
| Use Gradient Descent with a fixed step size | Matching Pennies · Matching Pennies, η=1, initial conditions y₁⁰=y₂⁰=(1,0) | player 1 regret: \Theta(\sqrt{T}) | Mixed |
| Use Gradient Descent with a fixed step size | Online learning in 2x2 zero-sum games with a unique fully mixed Nash equilibrium · 2x2 zero-sum games | time-average strategy convergence | Helped |
Closest papers by meaning
- No-regret Learning in Harmonic Games: Extrapolation in the Face of Conflicting Interests (NeurIPS 2024 · similarity 0.93)
- Optimism Without Regularization: Constant Regret in Zero-Sum Games (NeurIPS 2025 · similarity 0.93)
- On the Convergence of No-Regret Learning Dynamics in Time-Varying Games (NeurIPS 2023 · similarity 0.93)
- Convergence of No-Swap-Regret Dynamics in Self-Play (NeurIPS 2024 · similarity 0.92)
- Best of Both Worlds: Regret Minimization versus Minimax Play (ICML 2025 · similarity 0.91)
- No Coin Left Behind: Maximizing Strategic Surplus Against No-Regret Dynamics (NeurIPS 2026 · similarity 0.91)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.